Counseling is an effective method for treating depression but little is known about the content of counseling and interpersonal care with depressed patients in primary care settings. The information we have comes from structured protocol studies of counseling under highly controlled conditions within the mental health specialty sector. It is important to better understand how primary care clinicians usually counsel their depressed patients in general medical settings because depression is common in this treatment sector; within this setting only half of depressed patients receive adequate treatment, and under managed care many depressed patients will only receive care for mental health problems within the general medical sector. Systematic data describing the content of usual care counseling will also help guide physician reimbursement policy because counseling is exactly the kind of care that depressed patients are paying for when they purchase healthcare insurance. Moreover, it is this kind of care that will affect the costs and outcomes of care among depressed patients. This FIRST proposal will build upon the Principal Investigator's previous work on clinician usual counseling and interpersonal style of care for depression. Utilizing the data from two large "parent" studies of depression (AHCPR PORT-II, Kenneth B. Wells, PI and NlMH Cooperative Agreement, Lisa V. Rubenstein, PI), the proposed project will examine the content of usual counseling and interpersonal care provided to depressed patients by primary care clinicians. The PI for this FIRST study is a key investigator in these parent studies which are now in their first year of planning. Both parent studies are designed to improve the quality and cost-effectiveness of care for depression in prepaid group practices in six US sites. In this FIRST we will analyze the "black box" of counseling and interpersonal style in more detail than has been possible in prior large-scale health services research studies. We will conduct new analyses of the counseling data from 500 primary care clinicians and 2,500 depressed patients in the parent studies. New data will be collected through audiotaped content analysis of verbal counseling behavior from 225 encounters with depressed and non-depressed patients and their providers in one site. Analyses will provide a richer understanding of counseling and interpersonal style than has been previously available for the study of depression in primary care.